SemISAC realizes semantic communication and semantic sensing within a single dual‑function waveform. At the transmitter a joint semantic encoder maps a road‑scene image to semantic symbols that occupy the data cells of an OFDM grid, while the remaining cells serve as pilots for channel state estimation and sensing. The pilot density and placement are adaptively optimized according to the prevailing channel conditions to balance communication error rate and sensing accuracy. At the receiver a deep‑learning model reconstructs pixel‑wise road segmentation from the received waveform; simultaneously the transmitting vehicle captures reflected waveforms from surrounding objects and employs task‑specific decoders for target recognition and range estimation. Simulations are conducted in a vehicular scenario where vehicles share road segmentation and perform object classification and distance measurement. Results show that SemISAC achieves segmentation accuracy comparable to a dedicated semantic communication module, while markedly outperforming conventional and semantic baselines in target recognition and range estimation.
Review: This study demonstrates that adaptive pilot configuration can jointly optimize communication and sensing, offering a practical pathway for efficient 6G ISAC deployments.